基于机器视觉的铝铸件表面缺陷自动识别技术  

Automatic Identification Technology for Surface Defects of Aluminum Castings Based on Machine Vision

作　　者：郑晓玲[1] ZHENG Xiaoling(College of Mechatronical Engineering and Automation,Liming Vocational University,Quanzhou 362000,China)

机构地区：[1]黎明职业大学机电工程与自动化学院,福建泉州362000

出　　处：《黎明职业大学学报》2018年第3期78-83,共6页Journal of LiMing Vocational University

摘　　要：铝铸件在铸造过程中表面会出现裂纹、气孔等缺陷,影响其各项性能。依托先进的机器视觉技术,通过采用边缘检测算法检测出缺陷区域并将其分割,提取各缺陷的特征信息并进行描述和分析,并利用SVM (支持向量机)分类器对输入的特征信息进行分析并处理。研究表明:该算法能够实现智能分类识别铝铸件缺陷。Porosities and cracks are common defects in the process of casting aluminum castings, which will affect the performance of aluminum castings. Relying on the advanced machine vision technology, defect areas can be detected and segmented by using edge detection algorithm. Thus the characteristic information of each defect can be extracted, described, analyzed and finally be processed with a SVM classifier. The research shows that this algorithm can realize intelligent classification of aluminum casting defects.

关 键 词：铝铸件 缺陷识别 机器视觉 支持向量机 

分 类 号：TH164[机械工程—机械制造及自动化]